Process Control of Ionic Contamination Achieving 6-Sigma Criteria in The Assembly of Electronic Circuits
BY P. ECKOLD, L. HENNEKEN, R. FRITSCH, U. WELZEL OF ROBERT BOSCH GMBH, AUTOMOTIVE ELECTRONICS, STUTTGART AND M. ROUTLEY, G. NAISBITT OF GERMANY GEN3 SYSTEMS LIMITED, FARNBOROUGH, UK
Ionic contamination testing as a process control tool a newly developed testing protocol based on IPC-TM 650 2.3.25, was established to enable monitoring of ionic contamination within series production.
The testing procedure was successfully implemented within the production of high reliability, safety critical electronic circuits, involving multiple production sites around the world. I will be shown in this paper that the test protocol is capable for meeting Six-Sigma-Criteria.
For a Gauge R&R study, a calibration solution of 0.1 wt.-% NaCl was used in order to investigate the repeatability and reproducibility of the test protocol employing newly developed contamination testing systems, which were placed at five locations worldwide. A total failure range of below 0.1 μg/cm2 NaCl at a target value of 1.0 μg/cm2 NaCl (± 8.8%) was achieved, combined with manual laboratory handling of fluids (pipette, temperature).
For process control this value is acceptable and demonstrates that ionic contamination testing based on IPC-TM 650 2.3.25 is able to be used as a process control tool in manufacturing of electronic control units.
The increasing use of electronic control units in harsh environments, in, especially, high reliability, safety-critical applications, demands full understanding and rigorous control of electrochemical reliability.1-3
The production of electronic circuits involves more than 20 individual chemical process steps, most of which contain high levels of ionic matter. Electric potentials when combined with a damp or moist operating environment, coupled with the presence of ionic matter, can often increase the probability for dendritic, electrochemical reactions and premature, unpredictable circuit failures.4-6
For these reasons the analysis of the surface insulation resistance (SIR) can be employed to assess electrochemical reliability of assemblies and thereby establish a qualified manufacturing process.7
Electrochemical migration (ECM) is known as the migration of electrochemically mobile species between conductive paths caused by an applied voltage. Consequently, metal can dissolve at the anode and deposit at the cathode which leads to the growth of a metallic dendrite.2 Figure 1 (left) illustrates the influence factors for ECM. The combination of a closed water film and an applied voltage between two conductive paths is the necessity prerequisite for this failure mode.
The production of electronic circuits involves more than 20 individual chemical process steps, most of which contain high levels of ionic matter.
Ionic contamination on the surface of printed circuit boards (PCBs) is an acceleration factor which can:
- reduce the dissolution enthalpy of metals.
- reduce the dew point.
- shift the pH value towards higher /lower regions.
- increase the conductivity of the water film.
Thus, ionic contamination is not essential but a supporting factor for ECM. Therefore, the level of ionic contamination has to be controlled during the manufacturing of electronic assemblies using an appropriate process control tool in order to ensure that the individual production process steps (e.g. surface-mount or through-hole technology) are in control in view of the overall level of ionic contamination.
Validation of humidity robust design elements
The electrochemical reliability of design elements used in printed circuit board assemblies can be evaluated based on IPC- 9202 under defined climatic conditions. For this purpose the surface insulation resistance (SIR) has to assessed at design elements which pertain to actual series production conditions as closely as possible (e.g. series PCB materials, solder paste, …) applying the B52-test board. Applying specific climatic conditions the humidity robustness of design elements produced using specific material combinations can be evaluated using SIR measurement technology.
Figure 2 illustrates the principle layout of the B52 board proposed by IPC-9202. The corresponding design elements can be adapted according to the advanced product requirements regarding mi niaturization and component portfolio. Consequently, a design element specific test has to be performed in order to ensure the electrochemical reliability of PCBAs under high temperature and high humidity conditions.
Ionic Contamination Testing as a Process Control Tool
Currently used methods for detection of ionic contamination are described within IPC-TM 650 220.127.116.11 The original resistivity of solvent extract (ROSE) test was intentionally designed for process control of the cleaning process within manufacturing of electronic control units. Currently applied measurement protocols can be divided into two principal modes of operation: static and dynamic extraction. The different analyser designs combined with the variety of slightly varying measurements procedures makes it almost impossible to compare data from different equipment types. In addition, measurement data are often used focussing on an absolute cleanliness level of no-clean assemblies. It has to be pointed out that these measurement technologies were not designed for creating cleanliness data of electronic control units manufactured with no-clean processes.
However, ionic contamination testing is able to be used as a process control tool in order to monitor the level of ionic residues on electronic assemblies caused by several production processes (e.g. reflow, selective soldering …) over production time. For this purpose an optimized testing system and protocol based on IPC-TM 650 2.3.25 was implemented in collaboration with system suppliers in order to meet a Six-Sigma criteria, and also between several production sites.
PICT – process ionic contamination testing
Process ionic contamination testing (PICT) is an optimized testing process compared to conventional ROSE test procedures. PICT is applying a closed loop / static system according to the classification of IPC-TM 650 2.3.25. The working principle is illustrated in Figure 3. Test and regeneration cycles are independent from one another and dissolved ionic residues from the specimens are trapped into an ion exchange column during regeneration until the conductivity of the extraction solution is reduced to a specified level.
The flow rate of the solution was optimized to avoid undue turbulence during the test cycle. Thus, a fast removal of ionic contaminants from the printed circuit assembly is possible and the testing time can be reduced. Furthermore, the influence of the atmospheric reaction of carbon dioxide to carbonic acid was decreased. Nevertheless, the carbon dioxide compensation is a key factor in achieving reproducible measurement results for ionic contamination analyses. Dependent on the environmental conditions (e.g. atmospheric composition, temperature, relative humidity …) the solvent conductivity in increased due to the formation of carbonic acid and its subsequent decomposition to hydronium and carbonate ions. Test measurements with the testing solution under the corresponding environmental conditions and the application of a mathematical algorithm lead to a site-specific carbon dioxide tolerance factor for compensation. The analysis of solutions with low concentrations of ionic matter requires a high accuracy of the conductivity measurement cell. For this purpose a solid-state measuring cell is applied, which is connected to a ballistic amplifier. Thus, a measurement accuracy of ±0.005 μS/cm can be achieved.
In order to use ionic contamination as a process control tool in electronic manufacturing a gauge study was performed for analysing the repeatability and the reproducibility over different production sites. The test equipment was installed in five different production sites worldwide and set up according to the technical requirements of the supplier. In general, 2 ml of a test solution of 0.1 wt.-% sodium chloride was injected and circulated through the conductivity probe in the analyser. The constantly measured solution conductivity was recalculated to sodium chloride equivalent (NaCleq. / μg/cm2) based on a given surface area.
The implementation of ionic contamination testing as process control tool (PICT) exhibits advantages to current ROSE test protocols.
The results of the analysis on repeatability is shown in Figure 4. For this purpose, 40 measurements of a 0.1 wt.-% sodium chloride solution was carried out over three days at one manufacturing site. No outlier in the data was detected. Based on the existing data an internal failure of the analyser would have been detected with a probability of 96.8 %. Furthermore, a slight increase of the instruments reading was observed during each of the three days, with the level dropping overnight. This effect can be attributed to the temperature of the measurement solution during subsequently performed measurements during one day. An increased temperature of the solution leads to an increased mobility of ionic species and consequently to an increased conductivity level. The process was evaluated to be capable for allowing demand of tolerances (±10 %) achieving a Cgk value of 2.23. In total the repeatability of the analyser meets Six-Sigma-requirements.
A reproducibility study was carried out using five different analysers, located at different production sites worldwide. Using a set of unified work place instructions and directions regarding the local calibration procedure, an aligned test procedure was ensured. Especially consistency in the laboratory handling of fluids (e.g. pipette, temperature …) has to be taken into consideration when analysing the results. The total failure range was calculated to be ± 8.8 % which is an acceptable value for chemical analysis. For the charts in Figure 5 just slight deviation of the measurement values achieved in one site can be observed. Summarized, ionic contamination testing with optimized parameters is able to be used as a process control tool detecting ionic contamination on printed circuit board assemblies.
Influence of process parameters and process optimization
The previously described testing protocol can be used to study the influence of certain process parameters on the level of ionic contamination. Figure 6 illustrates a selection of influence factors on the final NaCleq. value on product level.
Variations in materials (e.g. solder paste, flux for selective soldering …) are strongly affecting the absolute value for ionic matter. Depending on the chemical activity and the encapsulation of flux residues after soldering, the NaCleq. value is varying. It has to be pointed out, that the scattering of the absolute value is increased on product level caused by numerous process steps during the assembly. However, it is possible to detect process deviations and corresponding root cause analysis focussing on process stability can be started and monitored.
Consequently, an overall limit of ionic contamination does not fit to the pool of materials and processes used in electronic manufacturing. Furthermore, ionic contamination testing has to be performed product specific as a process control tool.
As previously emphasised, electrochemical migration is caused by an electric field and a closed water film between neighbouring conductive paths. Ionic contamination is given by materials and processes and can accelerate the electrochemical processes under high temperature and high humidity conditions. Consequently, the measurement of ionic contamination does not predict the humidity reliability of printed board assemblies. The humidity robustness of design elements can be ensured by SIR measurement technology according to the principle of IPC 9202.
The implementation of ionic contamination testing as process control tool (PICT) exhibits advantages to current ROSE test protocols. Due to system optimization and the implementation of unified work place instructions, it was possible to achieve 6σ criteria for the repeatability and a sufficient reproducibility of recorded ionic contamination data at five different production sites worldwide. Using this approach, process control and methods of process optimizations are possible.
The authors would like to acknowledge their respective companies for providing financial support for this research.
1. D. Minzari, M.S. Jellesen, P. Moller, R. Ambat, Corros. Sci. 2011, 53, 3366-3379.
2. P.-E. Tegehall, Impact of Humidity and Contamination on Surface Insulation Resistance and Electrochemical Migration, Springer Verlag London, New York 2011, 227-253.
3. B.-I. Noh, S.-B. Jung, Circuit World 2008, 34, 8-13.
4. V. Verdingovas, M.S. Jellesen, R. Ambat, Corros. Eng. Sci. Tech. 2013, 48, 426-435.
5. D. Mizari, M.S. Jellesen, P. Moller, P. Wahlberg, R. Ambat, IEEE Trans. Dev. Mater. Reliab. 2009, 9, 392-402.
6. B.-I. Noh, S.-B. Jung, J. Mater. Sci.: Mater. Electron. 2008, 19, 952-956.
7. D.Q. Yu, W. Jillek, E. Schmitt, J. Mater. Sci.: Mater. Electron. 2006, 17, 229-241.
8. IPC-9201A – Surface Insulation Resistance Handbook, 2007.
9. IPC-9202 – Material and Process Characterization/Qualification Test Protocol for Assessing Electrochemical Performance, 2011.
10. IPC-TM-650 2.3.25. – Detection and Measurement of Ionizable Surface Contaminants by Resistivity of Solvent Extract (ROSE), 2012.
This paper was first presented at IPC APEX Expo 2017 in San Diego, CA, USA in February 2017